Agent Skills Hub

claude-code-mechanism-selector

A decision-support tool for Claude Code users to select the optimal extension mechanism—slash commands, skills, subagents, or hooks—based on project requirements.

llmintegration

Multi-model LLM integration patterns for Claude, GPT, Gemini, and Ollama. Features API handling, prompt engineering, token management, and model-agnostic orchestration.

looped-builder

Scaffold complex, multi-step coding tasks into actionable implementation plans and execute them autonomously using a Claude-driven bash loop.

langchain-chat-models

A unified interface for integrating and managing LLM chat providers like OpenAI, Anthropic, Google, Azure, and Bedrock within LangChain applications.

claude

Anthropic Claude AI models for high-performance coding, large-context analysis, and GUI interaction.

maestro

Intelligent RAG-based gateway that routes coding tasks to specialized Swift/iOS expertise without context window bloat. Uses MCP to retrieve precise patterns from 100+ indexed skills.

Create New Skills

A framework for creating reusable Claude Code agent skills, following best practices for directory structure, progressive disclosure, and multi-file patterns.

token-surgeon

Audit and optimize your AI prompts with Token Surgeon. Detect 10 common waste patterns, calculate efficiency, and reduce token usage for better prompt performance.

qras

A local RAG semantic memory system using Qdrant and Ollama. Ideal for recalling workspace files, notes, project decisions, and user preferences with high-relevance vector search.

context-bar

A color-coded, real-time context usage progress bar for the Claude Code statusline and manual on-demand checks.

context-compression

Optimize agent performance and token usage through advanced context compression, structured summarization, and task-oriented state management for long-running sessions.

querying-json

Efficiently extract, filter, and transform specific fields from JSON files using jq, saving up to 95% of context window usage compared to reading full files.